Current knowledge on the use of ultrasound measurements of fetal soft tissues for the assessment of pregnancy development

Ultrasound is very widely used in obstetrics. It enables detailed imaging of the fetus, among other possibilities. However, it is different at every stage of pregnancy. The use of diagnostic ultrasound begins already at 4–5 weeks of gestation when it is possible to visualize the gestational sac in the uterine cavity. From approximately 5–6 weeks of gestation the embryo’s echo is visible⁽¹⁾. As the pregnancy develops, the number of assessed parameters and the level of difficulty of the ultrasound procedure increase. The information acquired using this method concerning the developing embryo allow for the prediction, diagnosis and therapy of certain fetal pathologies. The advantages of diagnostic ultrasound primarily include non-invasiveness and ease of the procedure⁽¹⁾. According to the latest reports, ultrasound scans performed during pregnancy are safe and do not affect fetal weight, premature labor risk, the child’s condition at birth or perinatal mortality⁽²⁾. Sonography is also associated with certain limitations, especially when the visualization of the fetus is difficult due to unfavorable technical conditions, the mother’s obesity, anterior placenta or oligohydramnios⁽³⁾. The primary aim of ultrasound scanning in pregnancy is to limit the risk of obstetric complications by early detection of abnormalities⁽¹⁾. These include fetal growth abnormalities such as both intrauterine growth restriction (IUGR) and macrosomia, which are more common in pregnancies complicated by diabetes, obesity, hypertension, nicotine addiction or fetal genetic defects, among other conditions⁽¹⁾.

Over the last few years a larger proportion of neonates with high birth weight for gestational age has been observed. This phenomenon is called macrosomia, which can be defined as fetal weight exceeding 4500 g regardless of the fetal age or exceeding the 90^th percentile for a given gestational age and population^(1,4). In pregnancies complicated by diabetes the estimated fetal weight exceeds 4200 g⁽⁴⁾. The prevalence of macrosomia in the general population is 6–14.5% and in women with diabetes 25–42%⁽⁵⁾. The risk factors for macrosomia, alongside maternal diabetes, include maternal obesity, post-term labor, a history of giving birth to macrosomic children, multiparity, advanced maternal age, male fetuses and hyperinsulinemia⁽⁴⁾. Macrosomia is associated with multiple complications, both for the mother and the fetus. For fetuses and neonates these are increased mortality, shoulder dystocia, perinatal trauma (including humerus and clavicle fracture and damage to the brachial plexus), low Apgar score, postpartum hypoglycemia, prolonged jaundice, systemic defects, respiratory problems, calcium deficiency and infections, among other conditions^(1,4,5). For mothers, macrosomia is associated with an increased risk of perinatal trauma such as pelvic diaphragm and anal sphincter damage as well as an increased risk of the delivery by cesarean section and of postpartum hemorrhage^(1,4,5). Early detection of macrosomia allows to avoid the above complications.

Another type of intrauterine growth abnormalities is intrauterine growth restriction. This term is assumed to refer to fetuses whose estimated body weight or abdominal circumference value is below the 10^th percentile⁽⁶⁾. The possible cause of intrauterine growth restriction is considered to be impaired circulation of the fetoplacental unit, i.e. placental insufficiency^(1,6). The etiopathogenesis of this condition is still unknown, while the mechanisms responsible for intrauterine growth restriction can be classified as fetal, maternal, placental and environmental factors⁽¹⁾. The most common factors include maternal age of > 40 years, the mother’s short stature, low body mass, arterial hypertension, diabetes with vascular complications, nicotine addiction, obesity, kidney diseases, autoimmune diseases, anemia and cholestasis disease of pregnancy^(1,6). Fetuses with suspected growth deficiency include those with constitutional or genetically-related low body weight as well as those with pathological conditions such as intrauterine growth restriction. Low birth weight represents an important problem in obstetrics since it is responsible for 69.6% of neonatal deaths and 66.4% of intrauterine deaths⁽⁷⁾. The best diagnostic method for this pathology is regular estimation of fetal weight or measurement of fetal abdominal circumference in high-risk pregnancies⁽⁶⁾.

Currently, morphometric formulae are used for estimating fetal weight. They utilize basic biometric parameters such as biparietal diameter, head circumference, abdominal circumference and femur length⁽⁸⁾. Hadlock and Shepard formulae are most commonly used; however, they have certain limitations. The sensitivity and specificity of Hadlock formula in the detection of fetal macrosomia are 62% and 93%, respectively, and for Shepard formula they are 21% and 99%, respectively⁽⁹⁾. For this reason, researchers all over the world have been looking for other sonographic parameters correlating with fetal weight, with a higher predictive value. The examined fetal parameters include liver dimensions, cerebellar diameter, kidney length, upper arm soft tissue thickness, femur volume and cheek-to-cheek diameter^(9–12).

Based on the reports above, numerous studies proposing new fetal weight calculation formulae have been produced^(16,19,20). Currently, fetal abdominal circumference is the strongest predictor of fetal weight⁽¹⁹⁾. However, it has a large risk of error, particularly when it is not possible to ensure optimal conditions for the measurement during the scan⁽¹⁹⁾. Obtaining a high-quality ultrasound image is not easy, even for an experienced examiner. Measurements obtained from lower-quality visualizations have a larger error rate. For this reason, Scioscia et al. developed a new formula for estimating fetal weight⁽¹⁶⁾. Based on mathematical analysis they obtained a formula utilizing FL and FTSTT. The advantages of this formula include ease of examination and the possibility to perform measurements when the head of the fetus is located low in the pelvis and is inaccessible to the examiner⁽¹⁶⁾. Scioscia and Hadlock formulae were considered satisfactory to a similar extent in terms of the ability to estimate the actual fetal weight, with an absolute mean error rate below 15% in over 90% of cases^(16,20). These results are corroborated by a study by Abuelghar et al., who first verified the utility of Scioscia formula and then modified it to compare the results obtained using both formulae⁽²⁰⁾. According to the authors, both formulae displayed a significant correlation with the neonate’s birth weight; however, the strength of the original Scioscia formula was higher: r² = 0.609 (p < 0.001) for the modified formula, r² = 0.957 (p < 0.001) for the original one⁽²⁰⁾. However, the results of a study by Barros et al. published in 2016 do not confirm this. They observed a weak correlation for weight estimated using the Scioscia formula, which is based only on FL and FTSTT⁽²¹⁾. They also found that using this formula leads to overestimation of larger fetuses and underestimation of smaller ones. The authors believe that the reason for the differences was a low number of macrosomic fetuses included in their study. In this study ultrasound scans were performed by resident physicians with different experience at the admissions department since the aim of the study was to check the use of the formula in such conditions. Kalantari et al. also worked on the development of new formulae for fetal weight estimation⁽¹⁹⁾. The study included various combinations of customarily used sonographic parameters and FTSTT. The formula in which AC, HC and FL were combined with FTSTT displayed the highest predictive strength (r = 0.77).

Apart from soft tissue, some more advanced and detailed measurements are taken such as those involving adipose and lean tissue. The results of such measurements are presented in a study by Bernstein et al. The measurement involved visualizing the longitudinal section of the long bones – the femur and humerus in this case, and rotating the transducer by 90 degrees in order to obtain the transverse section of the fetal bones⁽²²⁾. Adipose tissue thickness was the difference between the total subcutaneous tissue and lean tissue thickness, i.e. muscle and bone thickness. The authors noted that both these parameters display a unique growth profile. Due to the accelerated fetal growth in late pregnancy these measurements can be a more sensitive and specific marker of fetal growth abnormalities. They were also used by Larciprete et al. and Galan et al., among others, as a more precise marker of fetal body structure^(23,24).

In comparison with numerous studies on new sonographic measurements in physiological pregnancies and those complicated by macrosomia, only few publications concern FTSTT in pregnancies with IUGR. Balouet et al. conducted ultrasound scans on 232 fetuses, only 39 of which were estimated to be too small for gestational age (SGA) ⁽²⁵⁾. In this group of fetuses the sonographic assessment of thigh soft tissues had a high positive predictive value (74%) in relation to low birth weight with a sensitivity of 74% and specificity of 94%. A study by Larciprete et al. estimated adipose tissue and lean muscle tissue of the humerus and femur in addition to fetal abdominal and subscapular soft tissue measurement⁽²³⁾. The authors of the study noticed significantly lower subscapular, abdominal and humeral subcutaneous tissue values in fetuses with intrauterine growth restriction. It is also one of the few studies which included patients with arterial hypertension (9 out of 14 examined patients had hypertension and 2 out of 14 had preeclampsia). An interesting observation in this regard may be a study by Galan et al., who studied two populations of pregnant women living at different altitudes 1600 m apart in absolute terms⁽²⁴⁾. The authors noticed that abdominal, thigh and upper arm subcutaneous tissue thickness values were lower in the fetuses of mothers living in the region located higher. However, no differences in muscle tissue values were found between the two populations. A different opinion is presented by Hill et al., who studied 240 patients for fetal thigh, calf and abdominal subcutaneous tissue parameters. In 13 fetuses intrauterine growth retardation was demonstrated and in 38 high body weight was found. The authors concluded that sonographic parameters could not be treated as reliable predictors in either of the groups⁽²⁶⁾.

Alongside two-dimensional (2D) sonography researchers are trying to find ways of using three-dimensional ultrasound (3D) for determining fetal weight. The authors of one of the first studies on this subject were Chang et al., who created a fetal weight estimation formula based on the analysis of 3D measurements of the thigh. Their formula had a higher accuracy than the methods used to date which utilized 2D measurements. The absolute error rate of the authors’ formula was 5.9%⁽²⁷⁾. Similarly, Liang et al. performed 3D measurements of the upper arm which they used to estimate fetal weight with a higher accuracy than the standard methods⁽²⁸⁾. Both studies were preliminary reports and 3D ultrasound measurements constituted the sole variable in the formula used. Schild et al., in contrast, developed a formula based on three-dimensional measurements of the abdominal, thigh and upper arm volume of the fetus as well as a standard BPD 2D measurement⁽²⁹⁾. The new formula had a much lower mean error rate than the standard formula based on 2D measurements. Even more precise estimation of body mass in small fetuses (below 1600 g) was achieved by combining 3D measurements with other biometric parameters of the fetus such as FL, HC and BPD⁽³⁰⁾. The authors concluded that the use of 3D measurements for fetal weight estimation is more precise, but at the same time more time-consuming: the measurement of each limb lasted 10–15 minutes on average. In order to increase the utility of 3D ultrasound Lee et al. in their study aimed to reduce the time needed to perform the measurements to 10 seconds by using a commercial program⁽¹²⁾. A hybrid transducer was used for the examination, which allowed to obtain AC, FL and BPD from volume measurements. The authors concluded that combining AC and BPD measurements with upper arm or thigh volume is associated with more precision. The mean error of fetal weight measurement in relation to the actual fetal weight was 6.6%. In the case of the standard Hadlock formula it was 8.5%.

The correct determination of fetal weight is an important issue in obstetrics since it has a significant impact on the course and end of labor and delivery. Incorrect estimation of fetal weight can result in multiple, often dangerous complications both for the pregnant mother and the fetus. The currently used Hadlock formula for fetal weight estimation has an error rate of 20%, which may depend on the skills of the examiner, equipment base, conditions of the examination as well as the stage of pregnancy or labor. For this reason, new parameters that can be used to predict fetal weight are being reported. The aim of researchers around the world is to find a diagnostic parameter that has as small an error rate as possible, is quick to use and reproducible by different examiners. Fetal soft tissue thickness measurements, both two- and three-dimensional, may prove to be such a parameter. The current reports on the topic are not consistent; studies were often conducted on small groups of patients, which significantly limits the possible application of these measurements in the future. Ultrasound measurement of subcutaneous tissue thickness in various areas of the body may prove to be a strong predictor of fetal weight essential for sonographic assessment of pregnancy.